2x Qwen 3.5 on M1 Mac: 9B builds a bot, 0.8B runs it (advanced-stack.com)

A recent demonstration showcased the capabilities of Qwen 3.5 on an older M1 MacBook, illustrating how the 9B model effectively generated a functional Telegram bot. The bot was designed to forward user messages to a local LM Studio OpenAI-compatible server running the smaller 0.8B model. This setup highlights the potential for smaller teams to leverage powerful AI tools on accessible hardware, allowing users to perform coding tasks locally without exposing sensitive data to external servers. Despite the slower performance on the M1, the demonstration emphasized practical use cases for sensitive and offline workloads, showing that even on a six-year-old machine, implementing a local AI coding toolchain is feasible. The user was able to quickly create and iterate on a functioning bot with minimal setup, suggesting that as AI technologies continue to improve, the efficiency and usability on mainstream hardware will only enhance. This development is significant for the AI/ML community as it democratizes access to advanced coding tools, making them powerful yet manageable for individual developers and small teams looking to maintain privacy and control over their data.